Is the well running dry? A century of rainfall and streamflow patterns in South Africa’s largest protected area

Major concern has been expressed over the water quality and quantity of the perennial river systems of the Kruger National Park

By Dr Dave Thompson, SAEON Ndlovu Node and Ashley Lipsett, University of the Witwatersrand

Global climate change is not only associated with increasing temperatures, but also more frequent and prolonged periods of extreme climatic events such as flooding and drought.

Climate change aside, such extremes are already commonplace in southern and South Africa, where ecological systems - particularly savannas, endure not only the annual seasonality of dry winters and wet summers, but also the periodic drier-than-normal conditions associated with the El Niño Southern Oscillation. This temporary reorganisation of atmospheric conditions disrupts global weather patterns (Fig. 1) and affects natural and modified ecosystems and the services they provide.

El Niño events historically occurred every three to five years, although recent scientific evidence suggests increasing frequency. For example, of the 27 events occurring between 1899 and 2003, 11 occurred in the most recent three decades.

Figure 1. An El Niño event typically has large ripple effects on climate around the world, particularly in the southern hemisphere. Depending on geographic location, local conditions during El Niño can be hotter (indicated in red), drier (brown), wetter (green), or combinations thereof, than ‘normal’. For southern and eastern Africa, the climate during such periods is atypically hot and dry (yellow).
[http://www.vox.com/2014/11/8/7177709/el-nino-2014-forecast-weakening]

Climate change and the continued disruption of weather patterns therefore pose significant threats to, especially, the already water-stressed semi-arid and arid regions of the planet.

Southern Africa is considered one of the most susceptible regions to climate change impacts as projected further increases in temperature and evaporation, documented increases in the number of warm periods, altered rainfall regimes, and modelled increases in the frequency and severity of extreme climatic events interact to negatively influence hydrological cycling and ecosystem integrity.

Without rivers, the freshwater lifelines of arid and semi-arid regions, much local biodiversity would not exist. Consequently the protection of rivers and their associated ecosystems is a research and management priority. Faced with the ominous, multiple threats of climate change it is crucial that we better -

understand how river systems are changing;
plan for these changes; and
accommodate predicted change scenarios into river and protected area management policies.

Kruger National Park

South Africa’s largest conservation area, the Kruger National Park (KNP), protects a variety of flora and fauna, as well a range of diverse habitats. The high levels of biodiversity in KNP are supported by the park’s varied geology, temperature and rainfall seasonality and latitudinal gradients, fire, and permanent and short-lived water sources.

All but one of the perennial river systems in the park have their headwaters some 100 km or more outside of the park boundary. Research and monitoring have shown that water quality and quantity here are highly reduced by anthropogenic activities such as industry, agriculture and mining.

A 2013 pilot study carried out by WITS Honours student Ashley Lipsett revealed that KNP is directly affected by El Niño and experiences periods of drought lasting approximately 18 months every two to seven years. In contrast, non-drought periods have been typified, especially in more recent years, by devastating floods.

In her thesis Ashley highlighted a direct correlation between the El Niño cycle and ‘low flow’ of a subset of KNP rivers. However, scant additional work exists on the behaviour of KNP rivers as influenced by global climate change. One study (see Odiyo et al., 2007) links modelled river flow in the northernmost Luvuvhu River with rainfall variability, and another (see Kleynhans, 1996) highlights a trend towards annual, rather than perennial, flow dynamics for the same river.

A near 100-year window

To address this shortfall, Ashley - now an MSc student from the School of Geography, Archeology and Environmental Studies (University of the Witwatersrand, Johannesburg) - is currently conducting trend-analyses of daily rainfall and streamflow data records for much of the South African lowveld (Fig. 2). These records date back as far as 1920 in some instances, providing a near 100-year window through which to investigate hydrological responses in relation to a changing climate. The records were obtained from the South African Weather Service and the South African Department of Water and Sanitation respectively.

Figure 2. The current change-over-time analyses, dating back to the 1920s in some instances, include rainfall and streamflow records from multiple South African Weather Service weather stations (red pentagons) and Department of Water and Sanitation gauging stations (blue circles), located in river catchments within or to the west of the Kruger National Park

Rainfall records cover 30 weather stations located in river catchments partly, and rarely wholly, within KNP, with 22 river gauging stations sited within or immediately west of the protected area providing corresponding flow data (Fig 3).

Figure 3. Department of Water and Sanitation gauging weir B8H0016 provides real-time streamflow data for the perennial Tzendze river in central KNP (top). For much of the year, the annual Shisa River in the far north of KNP remains dry (bottom), here pictured upstream from gauging station B9H001. (Pictures: Ashley Lipsett)

Under direction from Prof. Stefan Grab (WITS) and Dr Dave Thompson (SAEON Ndlovu Node), and with support from SANParks colleagues, the research being carried out aims to establish and compare rainfall and streamflow patterns over time for the eight main river catchments of KNP, and to correlate changes in inter-annual rainfall and streamflow. Initial time-series analyses suggest a breakdown in the predictability of the El Niño cycle from the end of the 20^th century, and that climate extreme flood and drought events are becoming more severe (Fig. 4).

Figure 4. Plotting rainfall variability (positive wet years or negative dry years) on either side of the long-term annual average of 521.6 mm for southern KNP collectively has revealed that, since 1960, there has been a breakdown in the wet-dry seven-year cycle, and that extreme climate events, both flooding and drought, have increased in intensity

Typically, streamflow volumes have decreased substantially over time - particularly for the larger perennial rivers (Fig. 5), which in some cases have adopted an annual flow dynamic in recent years.

Figure 5. River flow in one of the Kruger National Park’s larger perennial rivers - the northern plains Shingwedzi, all but ceased in around 2000, despite variable regional rainfall. Since this date, the river has adopted a primarily annual flow dynamic, responding only to heavy local rainfall (top). On the other hand, seasonal systems, such as the Shisa in the far north of KNP, seem little impacted by time-trends since 1960, with flow reflecting local rainfall (bottom).

The ongoing research will further determine the effects and scale of impact of specific climatic phases, such as El Niño drought and tropical storm flooding, on streamflow. Because the river systems under study span the prevailing north-south rainfall gradient of north-eastern South Africa, all patterns and river behaviours are being interpreted within this spatial framework.

Determining the influence of climate on river dynamics

Comparing river systems that originate either within or beyond direct human influences, as enforced by the protected area boundary, will allow for a better understanding of how climate, in the absence of human land-use activities, influences river dynamics. Finally, and based on available rainfall projections, the researchers plan to model the likely future scenarios of rainfall and streamflow in KNP under plausible future climates.

Major concern has been expressed over the water quality and quantity of the perennial river systems of KNP, with the systems being significantly impacted by multiple land-use types before the water even reaches the western boundary of the conservation area. In line with the 1998 Water Act, park management currently focuses on water management at the catchment level to promote a balance between water use to sustain basic human needs and the sustainability of the water resource itself.

Crucial to this is the fundamental understanding that will emerge from this study regarding the interplay between river behaviour and a changing climate, and the relative impacts of climate and land-use on river flow.